A novel fault-tolerant control strategy for near space hypersonic vehicles via least squares support vector machine and backstepping method

Jia Song, Jiaming Lin, Erfu Yang

Research output: Contribution to conferencePaper

2 Citations (Scopus)
134 Downloads (Pure)

Abstract

Near Space Hypersonic Vehicle (NSHV) could play significant roles in both military and civilian applications. It may cause huge losses of both personnel and property when a fatal fault occurs. It is therefore paramount to conduct fault-tolerant research for NSHV and avoid some catastrophic events. Toward this end, this paper presents a novel fault-tolerant control strategy by using the LSSVM (Least Squares Support Vector Machine)-based inverse system and Backstepping method. The control system takes advantage of the superiority of the LSSVM in solving the problems with small samples, high dimensions and local minima. The inverse system is built with an improved LSSVM. The adaptive controller is designed via the Backstepping which has the unique capability in dealing with nonlinear control systems. Finally, the experiment results demonstrate that the proposed method performs well.
Original languageEnglish
Pages174-182
Number of pages9
DOIs
Publication statusPublished - 24 Oct 2016
Event22nd International Conference on Automation and Computing, ICAC 2016 - University of Essex, Colchester, United Kingdom
Duration: 7 Sep 20168 Sep 2016
http://www.cacsuk.co.uk/index.php/conferences

Conference

Conference22nd International Conference on Automation and Computing, ICAC 2016
Abbreviated titleICAC 2016
CountryUnited Kingdom
CityColchester
Period7/09/168/09/16
Internet address

Keywords

  • near space hypersonic vehicle
  • least squares support vector machine;
  • inverse system control
  • fault-tolerant control
  • fault tolerance
  • backstepping
  • control systems
  • attitude control
  • mathematical model
  • nonlinear control systems

Fingerprint Dive into the research topics of 'A novel fault-tolerant control strategy for near space hypersonic vehicles via least squares support vector machine and backstepping method'. Together they form a unique fingerprint.

Cite this